Median Compressor bad SMD solder

[comradehoser]

Here is my post transferred from general questions forum. This began because I wasn't detecting any noticeable effect, and now I am looking for probable causes to fix this:

Was testing [Median compressor] last night with the blend all the way CW, and a big crackle storm with jumps in volume started up in the pedal. I've seen this before in my builds, in a Deofol with SMD J201s--it was a poor SMD solder joint/connection issue. I am starting to really detest SMDs, as apparently I can't solder them worth a shite.

[During] the crackle bursts, when output was higher, I could hear what sounded like the compression working. So it sounds like a partial SMD connection is disabling part of the board. My money is on the 5457s, as the 1678s went onto the breakout boards like butter. Now I just have to find out which of the 38 different SMD solder points is faulty and remedy that without messing up SMD legs or lifting traces.

schematic:


If I'm understanding/interpreting correctly and correlating with electrosmash's analysis of the MXR Dyna-comp, the two transistors are involved in the envelope detector block (?) So if I am having crackles and on/off compression effect, it could be here, or in IC3 which also looks to be part of the envelope detector, or IC2, which looks like the OTA amplifier. IC 1 and 4 look unrelated to the actual effect implementation other than being buffers/final stage amplifiers (?)

 

[Gordo]

I absolutely torched the smd pads on a J201 on the Median and ended up getting another board. I don't usually have issues but sometimes the damage is too much as it was in my case.

Of course this doesn't help you at all but felt good to get it off my chest.

 

[comradehoser]

Here go the Vs, Comp maxed, level noon, tone noon, attack full CCW, release full CW:

Measured during crackle storming:

Q1
1: 1.730
2: 1.718
3: 1.148 variable

Q2
1: 8.95
2: 2.773 variable
3: 1.702 variable

IC1
1: 2.956
2: 2.975
3: 2.845
4: 0
5: 0.1
6: 4.490
7: 5.968
8: 8.95

IC2
1: 4.5
2: 4.46
3: 4.12
4: 0
5: 5.94
6: 5.936
7: 5.936
8: 8.95

IC3
1: 4.45
2: 4.49
3: 4.49
4: 0
5: 4.489
6: 4.49
7: 4.491
8: 8.95

IC4 (more variable than the others
1: 4.48
2: 4.49
3: 4.49
4: 0
5: 4.47
6: 4.49
7: 4.47
8: 8.95

 

[comradehoser]

Decided in my delirium to start an audio probe at midnight last night. Crackle storm is forward of Q2, pin 1, so I'm going to do some actions in that area--reflow solder on the daughter/adapter board, pull the pins and make sure there's no bridging under the IC daughterboard. In my infinite wisdom, I purchased 5 ICs, so if anything, I can just sub and see if that does the trick.

The one time I didn't use sockets, and I really wish I would've used sockets.

 

[comradehoser]

Apparently, it's not IC 2. And full CW blend results in no signal; not volume loss, volume absence. I don't think that's right.

Am going to pull, reflow, and socket daughter boards so I can at least try other ICs to see if it's my SMD soldering incompetence that's to blame.

At that point, if there's still an issue, I'll look at scouring off the 5457s and trying to find and install them in through hole.

 

[Locrian99]

Apparently, it's not IC 2. And full CW blend results in no signal; not volume loss, volume absence. I didn't think that's right.

Am going to pull, reflow, and socket daughter boards so I can at least try other ICs to see if it's my SMD soldering incompetence that's to blame.

At that point, if there's still an issue, I'll look at scouring off the 5457s and trying to find and install them in through hole.

If you need thru hole last i looked stomp box parts still had them, and last time i bought some from them they were specced very tight.

 

[mybud]

The one time I didn't use sockets, and I really wish I would've used sockets.

I fear I'm in the same boat as you but I'll post a separate thread so as not to hijack yours. Maybe we can share ideas/progress on this as we go?

Between us, I went the opposite way, starting with through hole 5457s. At the time it struck me that I should use sockets for the two JFETs but didn't, thinking in my optimism that it will work anyway.

Well, it very obviously doesn't and I think I know why. The key is this mystical IDSS thang, which is a constant for each and every JFET and highly variable so it's possible (it seems) to have the correct bias but the JFETs are so far out of spec that they won't turn off.

Yeah, it's counterintuitive but correct, I think. So if you're hearing signal from the get-go, it may be that your JFETs are letting too much signal through when they need some depletion to operate as variable resistors, what have you.

I had such high expectations for this one but you can't win 'em all. I'm keen to know how you get on but all in good time, I guess

 

[comradehoser]

I fear I'm in the same boat as you but I'll post a separate thread so as not to hijack yours. Maybe we can share ideas/progress on this as we go?

I am definitely down for some hijackery. At least I'll know that I'm not going crazy alone. Then we can be crazy together.

 

[mybud]

I am definitely down for some hijackery. At least I'll know that I'm not going crazy alone. Then we can be crazy together.

Sounds better than good to me.

 

[comradehoser]

Well, I'm sorry to say, this is my very second build where the troubleshooting stops here and will never be marked solved.

Desoldering the 1678 boards and lifted traces. I just can't deal with that and whatever other things are up with the build.

Adieu, monsieur.

Guess I will get around to trying again with another board.



Goddamit.